Mistral AI has released Voxtral 3, a speech model family pairing three transcription models with a text-to-speech model and a voice agent SDK built on LiveKit. Voxtral Small Transcribe 3 records a 2.1% word error rate and ranks third on Artificial Analysis, with all four models available in public preview on the Mistral API. https://www.marktechpost.com/2026/10/20/mistral-voxtral-3-speech-models-voice-agent-sdk/ #AIagent #AI #GenAI #AgenticAI
MarkTechPostMistral AI Releases Voxtral 3: Speech-to-Text, TTS and a Voice Agent SDKMistral Voxtral 3 adds speech-to-text, text-to-speech and a voice agent SDK, with transcription from $0.0035 per audio minute.